Clinical benefit of luspatercept treatment (tx) in transfusion-dependent (TD), erythropoiesis-stimulating agent (ESA)–naive patients (pts) with very low-, low- or intermediate-risk myelodysplastic syndromes (MDS) in the COMMANDS trial.

Abstract

6565 Background: There is an unmet need for effective tx that provides durable benefit for pts with anemia due to lower-risk MDS (LR-MDS). Here we report clinically meaningful responses to luspatercept tx in TD, ESA-naive pts with LR-MDS in the COMMANDS trial. Methods: Eligible pts were ≥ 18 years of age, had LR-MDS with or without ring sideroblasts and < 5% bone marrow blasts, endogenous serum erythropoietin < 500 U/L, required red blood cell (RBC) transfusions (defined as 2−6 RBC units/8 weeks [wk] for ≥ 8 wk prior to randomization), and were ESA-naive. Pts were randomized 1:1 to subcutaneous administration of luspatercept (1.0–1.75 mg/kg) once every 3 wk or epoetin alfa (EA; 450–1050 IU/kg) once weekly for ≥ 24 wk. New assessments of clinical benefit reported here include achievement and duration of ≥ 50% reduction in RBC units transfused over ≥ 12 wk (wk 1–end of tx [EOT]), transfusion burden (TB) on tx (wk 1–24), time to first transfusion, achievement and cumulative duration of all separate RBC transfusion independence (RBC-TI) ≥ 12 wk response episodes, and mean hemoglobin (Hb) increase ≥ 1.5 g/dL over wk 1–24. Results: As of March 31, 2023, 151/182 (83.0%) luspatercept- and 121/181 (66.9%) EA-treated pts achieved ≥ 50% reduction in RBC units transfused over ≥ 12 wk (wk 1–EOT; P= 0.0002), with median (95% confidence interval [CI]) durations of 130.0 (120.6–not evaluable [NE]) and 77.0 (54.9–123.1) wk, respectively ( P = 0.0004). A greater proportion of luspatercept vs EA pts achieved ≥ 50% reduction in RBC units transfused over ≥ 12 wk (wk 1–EOT), regardless of baseline (BL) TB: 105/118 (89.0%) luspatercept vs 82/111 (73.9%) EA pts with BL TB < 4 RBC units/8 wk and 46/64 (71.9%) luspatercept vs 39/70 (55.7%) EA pts with TB ≥ 4 RBC units/8 wk. The median (interquartile range) number of RBC units transfused during wk 1–24 of tx was 1.0 (0–5.0) in the luspatercept arm and 3.0 (0–8.0) in the EA arm. The median (95% CI) time to first transfusion was 155.0 (80.0–266.0) days for luspatercept vs 42.0 (23.0–55.0) days for EA pts ( P < 0.0001). Among pts who achieved RBC-TI ≥ 12 wk (wk 1–24), 22/124 (17.7%) luspatercept pts vs 12/88 (13.6%) EA pts achieved ≥ 2 separate RBC-TI ≥ 12 wk response episodes and cumulative median (95% CI) duration of all response episodes was 147.9 (122.0–NE) wk in the luspatercept arm and 95.1 (73.1–NE) wk in the EA arm ( P = 0.0067). Mean Hb increase ≥ 1.5 g/dL over wk 1–24 was achieved by 135/182 (74.2%) luspatercept pts and 95/181 (52.5%) EA pts ( P < 0.0001). Conclusions: Significantly greater proportions of luspatercept vs EA pts achieved improvements in Hb levels, reduction in TB and RBC units transfused, and had durable RBC-TI responses. Luspatercept provided clinically meaningful outcomes, supporting its use as the preferred tx for ESA-naive pts with LR-MDS-associated anemia. Clinical trial information: NCT03682536 .